I have wanted to build a decent subwoofer for a while that will go way down deep to compliment my fullrange speakers. Nothing loud or wall vibrating, just clean and low.
I saw some Peerless xxls 830843 drivers on ebay, found the specs, put them into Winisd.....perfect. Or so I thought! They are specified for a 'sealed' alignment which was ideal and gave me the excuse to use a couple of Rob Cheng Linkwitz Transform kits that I have to keep the box size reasonable and assist the drivers digging deep.
I bought them and tested them using the Dayton Audio Woofer tester, fully calibrated, but found that the measured specs, especially the Fs were way off. Now Winisd gves them as 'Ported' although obviously 'Sealed' is stll a viable option as they are borderline.
Question is, which specification do I believe for the purposes of final design? Is the Woofer Tester reliable or does it have issues? I did a full check to make sure the tester was set up properly on my computer before doing the test and that the leads and resistor calibration test was done.
Peerless Specs: Measured Specs:
Qts: 0.448 0.451
Qes: 0.471 0.478
Qms: 9.03 7.931
Fs: 25 30.28
Bl: 12.5 12.94
Spl: 86.74 87.5
Vas: 86.7 66.72
I'm torn between an Isobarik EBS in a 68 litre box (-3db @ 20Hz) and two 'sealed' Linkwitz Transformed boxes. The latter would have been better with the Peerless specs however as the Fs is given as 5Hz lower.
Any thought most welcome.
Cheers
I saw some Peerless xxls 830843 drivers on ebay, found the specs, put them into Winisd.....perfect. Or so I thought! They are specified for a 'sealed' alignment which was ideal and gave me the excuse to use a couple of Rob Cheng Linkwitz Transform kits that I have to keep the box size reasonable and assist the drivers digging deep.
I bought them and tested them using the Dayton Audio Woofer tester, fully calibrated, but found that the measured specs, especially the Fs were way off. Now Winisd gves them as 'Ported' although obviously 'Sealed' is stll a viable option as they are borderline.
Question is, which specification do I believe for the purposes of final design? Is the Woofer Tester reliable or does it have issues? I did a full check to make sure the tester was set up properly on my computer before doing the test and that the leads and resistor calibration test was done.
Peerless Specs: Measured Specs:
Qts: 0.448 0.451
Qes: 0.471 0.478
Qms: 9.03 7.931
Fs: 25 30.28
Bl: 12.5 12.94
Spl: 86.74 87.5
Vas: 86.7 66.72
I'm torn between an Isobarik EBS in a 68 litre box (-3db @ 20Hz) and two 'sealed' Linkwitz Transformed boxes. The latter would have been better with the Peerless specs however as the Fs is given as 5Hz lower.
Any thought most welcome.
Cheers
Both Peerless specs and your measurement look OK.
It's no surprise that the suspensions of a new driver are stiffer than specified by manufacturer's datasheet, giving mainly lower Cms, lower Vas, and higher Fs.
Measured stiffness also depend of voltage of the signal used for measurement: it's not a constant, nor are the derived parameters, Vas, Fs, etc...
Btw loudspeakers are not high precision devices, nor are the cabinets we calculate.
You must allow for tolerances and don't get mad trying tooptimize things that you can't by their own nature...
...
It's no surprise that the suspensions of a new driver are stiffer than specified by manufacturer's datasheet, giving mainly lower Cms, lower Vas, and higher Fs.
Measured stiffness also depend of voltage of the signal used for measurement: it's not a constant, nor are the derived parameters, Vas, Fs, etc...
Btw loudspeakers are not high precision devices, nor are the cabinets we calculate.
You must allow for tolerances and don't get mad trying tooptimize things that you can't by their own nature...
...
Last edited:
I agree, but it might be an idea if a company as large as Peerless could give an initial T/S spec. sheet with the expected changes when run-in.
I'm certainly not gettng mad about it, far from it. These drivers will deliver what I'm looking for either way.
I must make a decision which way to go sooner or later however, so any recommendations would be well considered. I'm after taught very deep bass but not at any great volume as it will be for music only.
Cheers
I'm certainly not gettng mad about it, far from it. These drivers will deliver what I'm looking for either way.
I must make a decision which way to go sooner or later however, so any recommendations would be well considered. I'm after taught very deep bass but not at any great volume as it will be for music only.
Cheers
I think providing two sets of parameters is very likely to increase confusion, and in the long term the only relevant ts parameters are those for a broken in driver.
These values are probably the mean of a significant number of samples. Significant variance is to be expected.
Your correlation with the manufacturer's given measurements is actually very good to excellent given that low/no hour drivers were tested.
Qms will increase, and Fs decrease with break in, overall though the Qts is pretty close to the stated value despite the low initial Qms.
In this case I would either break them in and remeasure or based on the good correlation of most measured parameters assume the manufacturer is competent and use their params to design the box.
These values are probably the mean of a significant number of samples. Significant variance is to be expected.
Your correlation with the manufacturer's given measurements is actually very good to excellent given that low/no hour drivers were tested.
Qms will increase, and Fs decrease with break in, overall though the Qts is pretty close to the stated value despite the low initial Qms.
In this case I would either break them in and remeasure or based on the good correlation of most measured parameters assume the manufacturer is competent and use their params to design the box.
corrected specs from peerless spec sheet in bold. Typical batch-batch variations are perhaps 5% on many items, but up to ~40% on Vas and ~20% on Fs. Compliance variations typically have little effect on frequency response, this one is roughly -2dB below spec at <60Hz in a 50L sealed box
I think you would find closer agreement if you added a few grams of weight to the cone, but before altering the driver run it in for a while.
Here is a comparison of specs, you can see that if your measurements are to be believed, the cone may be a bit light.
Woofer parameters
Fs 25 30.3 Hz
Qts 0.447 0.451
Qes 0.470 0.478
Vas 86.7 66.7 Liters
Re 5.8 5.8 Ohms
Dia 21.18 21.18 cm
Xmax 12.5 12.5 mm
Woofer Calculated parameters
Sd 0.03523 0.03523 m^2
Cms 0.000497 0.000383 m/N
Mdt 0.0815 0.0721 kg
Bl 12.57 12.91 T*m
no 87.92 89.22 dB/2.83V
no 86.53 87.82 dB/1W
Qms 9.031 7.984
Cone Sag 0.40 0.27 mm
Last edited:
Assigning a set of fixed parameters to an inherently nonlinear device is also fraught with pitfalls. Any difference less than 20% is pretty much noise, IMO.
Try doing an uncertainty analysis on T/S parameter calculation, it isn't pretty, let me tell you...
I was meticulous in following the given procedure whilst using the Woofer Tester and I did the test a few times to double check, and I do realise that it is hardly an industry standard device for state of the art measurement so I can therefore assume that the measurements that I took can be used with the caveat that a break-in period is obligatory.
My main concern initially was the difference in Fs, as I had modelled on Winisd Pro for a sealed box box using an LT. I was of the belief that the measurements needed to be reasnably accurate for an LT
So a ported box for a couple of months to break-in and then an accurate measurement of Fsc for an LT to get the bass I want, might be a good way to go!?
Cheers
My main concern initially was the difference in Fs, as I had modelled on Winisd Pro for a sealed box box using an LT. I was of the belief that the measurements needed to be reasnably accurate for an LT
So a ported box for a couple of months to break-in and then an accurate measurement of Fsc for an LT to get the bass I want, might be a good way to go!?
Cheers
It doesn't matter how meticulous you were or whether your measurement device is 'industry standard' or not, when the T/S model does not describe the system in any better than a ~90% fashion.
Parameters like Cms, Bl, etc... vary with excursion and Rms with frequency and excursion - so a measurement at 1/4V and 1V can look like different drivers. Do you know what level the Mfg took their measurement? Do you know what level you took your measurement? How about temperature? Temperature has a large effect on Cms (Vas).
You basically are whining about normal parameter variations. Typical newbie mistake.
Are you planning a FAST set up?
What frequency is xover?
Have you considered this? Sub-Woofer Controller
It doesn't require xover as the LT does
Last edited:
'Whining' is a bit strong and completely unnecessary. Please be constructive or do not bother making a comment. Use your 'experience' in aiding people and not to belittle them.
I make observations that I have found with the experience that I have and with the equipment I have available.
What you say basically means that the variations due to manufacturing tolerances and measurement techniques and conditions, and the variations of 'my' measurement conditions, render measurements all but guesswork in all but a tightly controlled and fixed environment. This does not help at all!
I make observations that I have found with the experience that I have and with the equipment I have available.
What you say basically means that the variations due to manufacturing tolerances and measurement techniques and conditions, and the variations of 'my' measurement conditions, render measurements all but guesswork in all but a tightly controlled and fixed environment. This does not help at all!
Hi scottjoplin,
the subs will be for a FAST set up and the frequency that they will come in at will be around 40 Hz but yet to be finalised.
I have looked at the 'subwoofer controller' before as I do like the ESP site. It's about the best audio site on the web. I'll take another look, but I do have a pair of very nice Linkwitz Transform kits tha have a 20 Hz high pass filter built in if required and an on-board power supply. It would be nice to use if I need it but if not then I'll probably sell them on when I have a clear out of audio related stuff in the near future.
I'll check out the controller again though.
Cheers
the subs will be for a FAST set up and the frequency that they will come in at will be around 40 Hz but yet to be finalised.
I have looked at the 'subwoofer controller' before as I do like the ESP site. It's about the best audio site on the web. I'll take another look, but I do have a pair of very nice Linkwitz Transform kits tha have a 20 Hz high pass filter built in if required and an on-board power supply. It would be nice to use if I need it but if not then I'll probably sell them on when I have a clear out of audio related stuff in the near future.
I'll check out the controller again though.
Cheers
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.